Isocitrate Dehydrogenase from the Hyperthermophile Aeropyrum pernix: X-ray Structure Analysis of a Ternary Enzyme–Substrate Complex and Thermal Stability

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
 8
 
  Isocitrate Dehydrogenase from the Hyperthermophile Aeropyrum pernix: X-ray Structure Analysis of a Ternary Enzyme–Substrate Complex and Thermal Stability
Share
Transcript
  Isocitrate Dehydrogenase from the Hyperthermophile Aeropyrumpernix  :X-rayStructureAnalysisofaTernaryEnzyme–Substrate Complex and Thermal Stability Mikael Karlstro¨m 1 *, Runar Stokke 2 , Ida Helene Steen 2 Nils-Ka˚re Birkeland 2 and Rudolf Ladenstein 1 1 Center for StructuralBiochemistry, Department of Biosciences at NovumKarolinska Institute, S-141 57 Huddinge, Sweden 2 Department of BiologyUniversity of Bergen, PO Box7800, Jahnebakken 5, N-5020Bergen, Norway Isocitrate dehydrogenase from  Aeropyrum pernix (  Ap IDH) isahomodimericenzyme that belongs to the b -decarboxylating dehydrogenase family and isthe most thermostable IDH identified. It catalyzes the NADP C and metal-dependent oxidative decarboxylation of isocitrate to  a -ketoglutarate.We have solved the crystal structures of a native  Ap IDH at 2.2 A˚, apseudo-native  Ap IDH at 2.1 A˚, and of   Ap IDH in complex with NADP C ,Ca 2 C and  d -isocitrate at 2.3 A˚. The pseudo-native  Ap IDH is in complexwith etheno-NADP C which was located at the surface instead of in theactive site revealing a novel adenine-nucleotide binding site in  Ap IDH. Thenative and the pseudo-native  Ap IDHs were found in an open confor-mation, whereas one of the subunits of the ternary complex was closedupon substrate binding. The closed subunit showed a domain rotation of 19 8  compared to the open subunit. The binding of isocitrate in the closedsubunit was identical with that of the binary complex of porcinemitochondrial IDH, whereas the binding of NADP C was similar to that of the ternary complex of IDH from  Escherichia coli . The reaction mechanism islikely to be conserved in the different IDHs. A proton relay chain involvingat least fivesolvent molecules, the5 0 -phosphategroupofthenicotinamide–ribose and a coupled lysine–tyrosine pair in the active site, is postulated asessential in both the initial and the final steps of the catalytic reaction of IDH.  Ap IDH was found to be highly homologous to the mesophilic IDHsand was subjected to a comparative analysis in order to find differencesthat could explain the large difference in thermostability. Mutationalstudies revealed that a disulfide bond at the N terminus and a seven-membered inter-domain ionic network at the surface are major determin-ants for the higher thermostability of   Ap IDH compared to  Ec IDH.Furthermore, the total number of ion pairs was dramatically higher in  Ap IDH compared to the mesophilic IDHs if a cutoff of 4.2 A˚was used.A calculated net charge of only C 1 compared to K 19 and K 25 in  Ec IDHand  Bs IDH, respectively, suggested a high degree of electrostaticoptimization,which isknowntobeanimportantdeterminantforincreasedthermostability. q 2004 Elsevier Ltd. All rights reserved. Keywords:  isocitrate dehydrogenase; thermal stability; disulfide; ionicnetworks; domain movements *Corresponding author Introduction Enzymes from hyperthermophiles are oftenhighly homologous to their mesophilic counter-parts and their catalytic mechanisms are usuallyidentical. They must be stable enough to withstanddenaturation at temperatures above 80  8 C andsimultaneously maintain the flexibility required 0022-2836/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. Abbreviations used: IDH, isocitrate dehydrogenase;IPMDH, isopropylmalate dehydrogenase;  Ap IDH,  Aeropyrum pernix  IDH;  Ec IDH,  Escherichia coli  IDH; Bs IDH,  Bacillus subtilis  IDH;  Tm IDH,  Thermotoga maritima IDH;  Af  IDH,  Archaeoglobus fulgidus  IDH;  Pf  IDH, Pyrococcus furiosus  IDH.E-mail address of the corresponding author:mikael.karlstrom@csb.ki.se doi:10.1016/j.jmb.2004.10.025  J. Mol. Biol.  (2005)  345 , 559–577  for enzymatic activity. There appears to be no singlemechanism or structural feature that is responsiblefor the high thermotolerance of hyperthermostableproteins. The major reason for this is the relativelysmall free energy difference between the folded andthe unfoldedstate of a proteinandthe complex wayin which the small number of weak forces, deter-mining protein stability, interplay with each other.Upon comparison with the mesophilic homologs,the most common determinants for hyperthermo-stabilityareinthefirstlineastatisticalprevalenceof ionic interactions at the protein surface, increasedformation of large ionic networks, electrostatic opti-misation and the reduction of repulsive charge–charge interactions. 1–3 Furthermore, a reduction of the hydrophobic accessible surface area, increasedhydrogen bonding and structural compactness has been observed. 3,4 The elucidation of the mechan-isms and structural determinants responsible forthe extreme thermostability of proteins fromhyperthermophiles is of importance for under-standing the question of why hyperthermophilicproteins show very high thermotolerance in spite of a relatively small gain of free stabilization energycompared to the mesophilic homologs. It isexpected that hyperthermostable enzymes willhave a great potential in biotechnological andindustrial applications in processes at elevatedtemperatures.Isocitrate dehydrogenase (IDH) is a metal-dependent (Mg 2 C or Mn 2 C ) enzyme that catalysesthe subsequent dehydrogenation and decarboxyla-tion of isocitrate to  a -ketoglutarate using NAD C or NADP C as cofactor. 5 a -Ketoglutarate is anintermediate in the citric acid cycle as well as theskeletal carbon source in the biosynthesis of certainaminoacidresidues.Thesepathwaysareamongthefirst to have evolved in the history of life. 6,7 Consequently, IDH is distributed broadly through-out the three domains of life, Archaea, Bacteria andEukarya, with diverse primary structures anddifferent oligomeric states. 8 The IDHs have been distinguished into threesubfamilies based on sequence comparisons. 8,9 Allof the archaeal and most of the bacterial IDHs aregrouped together into subfamily I, eukaryotichomodimeric IDHs and some bacterial IDHs con-stitute subfamily II, whereas eukaryotic hetero-oligomeric IDHs form a third subfamily. Themembers of subfamily II show very little sequenceidentity with those of subfamilies I and III. Thereare also monomeric IDHs that show no significantsequence similarity to the other subfamilies.The crystal structures of IDH from  E. coli 10 ( Ec IDH, PDB code 3ICD) and  Bacillus subtilis 11 ( Bs IDH, PDB code 1HQS) in su bfamily I and of porcine heart mitochondrial IDH 12 (denoted belowas porcine IDH, PDB code 1LWD) in subfamily IIare all NADP C -dependent homodimers and haverevealed a common fold, shared also by the crystalstructures of the NAD C -dependent isopropyl-malate dehydrogenases (IPMDH), which belong tothesamefamilyof  b -decarboxylatingdehydrogenasesusing a substrate that is structurally related toisocitrate. 13–15 The crystal structure of monomericIDH from  Azotobacter vinelandii  (PDB code 1ITW) 16 is also topologically related to homodimeric IDHs.Thisappearstohavebeenaccomplishedbyapartialgene duplication, which resulted in a pseudo 2-foldsymmetry in one domain corresponding to theothersubunitofthedimericIDHs.IDHandIPMDHshare a unique cofactor binding site that is differentfromthewell-knownRossmannfoldfoundinmanyother dehydrogenases. 10,13,17 Only a few amino acidresidues appear to be responsible for the discrimin-ation between NAD C and NADP C . 18–21 The reaction mechanism of  IDH has beenextensively studied in  Ec IDH. 22,23 In the proposedmechanism,aprotonisremovedfromthe a -hydroxylgroup of isocitrate. Subsequently, a hydride ion istransferredin a stereospecific way from the a -carbonatomofthesubstratetoC-4 ofthenicotinamideringofNADP C ,oxidizingisocitratetooxalosuccinate. 24,25 In a second step, the  b -carboxylate group of oxalosuccinate is lost as CO 2 , 26 and is replaced by aproton in a stereospecific way to form  a -ketogluta-rate. 27 During both transition states the negativecharge on the hydroxyl oxygen atom of isocitrate isstabilised by a magnesium ion. However, the initialproton abstraction mechanism as well as the finalproton donation are still matters of debate (seesection The active site).The regulation of the activity of the differentIDHs is diverse. The NAD C -dependent eukaryoticheterooligomeric IDHs are allosterically regulated by the activators AMP 28 or ADP, 29 whereas  Ec IDHis regulated by the IDH kinase/phosphatase-mediated phosphorylation of Ser113, which inacti-vates the enzyme by sterically hindering the binding of isocitrate and by electrostatic repul-sion. 30–33 Recently, a self-regulating mechanism of activity was postulated in human cytosolicNADP C -dependent IDH. 34 The structural homology across the differentspecies in this enzyme family makes them verysuitable for comparative studies. We have chosen tostudy IDH from the strictly aerobic hyperthermo-philic archaeon  Aeropyrum pernix  (  Ap IDH). Pre-viously, we have described  Ap IDH as the mostthermostable IDH characterized, with an apparentmelting temperature of 110  8 C. 8 It is a Mg 2 C - andNADP C -dependent dimer with identical subunitsof 47.9 kDa and belongs to subfamily I. A shortstructure notice of   Ap IDH was published recently 35 where a disulfide-bond at the N terminus and aseven-membered ion pair network at the surfacewere related to increased thermotolerance, but nostability data were given. The mechanism of regulation of   Ap IDH has remained unknown. Toour knowledge, no gene encoding IDH kinase/phosphatase has been found in the genome of   A. pernix . 36 In order to gain more information about thecatalytic mechanism and the large domain move-ments presumably involved in catalysis, we havesolved the structure of a ternary complex of   Ap IDH 560  Structure and Thermal Stability of   A. pernix  IDH   with Ca 2 C –isocitrate–NADP C . The structure wascompared with the structure of the ternary complexof   Ec IDH and with the binary complex of porcineIDH from subfamily II, which has a conservedisocitrate binding site. 12 It is likely that the reactionmechanisms are identical in  Ap IDH,  Ec IDH andporcine IDH, since the isocitrate binding sites areconserved both structurally and by sequence.In order to perform a more thorough analysis of the factors responsible for the high thermotoleranceof this enzyme, we have started a mutationalapproach based on our refined crystal structure of wild-type  Ap IDH. The structure is reportedtogether with apparent melting temperatures of the mutants and is primarily compared to thestructure of   Ec IDH but to some extent also to Bs IDH. Results and Discussion Quality and description of the models In this work, three different  Ap IDH structures arereported: a native  Ap IDH solved at 2.2 A˚(PDB code1XGV), a pseudo-native  Ap IDH solved at 2.1 A˚(1TYO) and a ternary complex solved at 2.3 A˚(1XKD). The ternary complex was formed in asoaking solution where citrate and Mg 2 C wereexchanged with  d -isocitrate, Ca 2 C and NADP C .The pseudo-native  Ap IDH was soaked with etheno-NADP C  but the etheno-NADP C molecule did not bind to the active site. However, the etheno-adeninemoiety of etheno-NADP C was located at thesurface of   Ap IDH, making contact also with asymmetry-related molecule, which resulted in aunit cell with the  c  axis elongated by 8.7 A˚. Thiscomplex is designated as “pseudo-native” becauseof the unusual binding of etheno-NADP C outsideof the active site and because it is basically identicalwith native  Ap IDH and has the same overallconformation. The 2 j F o j K j F c j  and  j F o j K j F c j  differ-ence density maps of native  Ap IDH and pseudo-native  Ap IDH contained uninterpretable density inthe isocitrate binding site. Most likely, the densityoriginated from citrate, since 0.1 M citrate waspresent in the crystallization buffer.  Ap IDH was crystallized as a dimer in theasymmetric unit with a solvent content of 52%corresponding to a Matthews coefficient of 2.6 A˚  3 Da K 1 . The two subunits are referred to assubunit A and subunit B. The space group was P 4 3 2 1 2. Table 1 summarizes the quality of thedifferent models. The quality of the B subunits issomewhat reduced in all models, since the densitiesof the large domain suffered from disorder, whichwasprobablycausedbythefewcrystalcontactsthisdomain is involved in. Accordingly, less solventmolecules were built into subunit B as compared tothe number in subunit A. Arg103 in subunit B of native  Ap IDH is found in a disallowed confor-mation, which can be explained by its location in asharp turn between helix c and strand C. In theternary complex, Ser120 in subunit A is also foundin a disallowed conformation; this is explained byits location at the end of a loop, which isdisordered. Overall fold  In contrast to many other hyperthermophilicenzymes,  Ap IDH is larger than its mesophilichomologs. Each subunit contains 435 residues,whereas  Ec IDH and  Bs IDH have only 416 and 423residues, respectively. In  Ap IDH, residues 1–131and 322–435 belong to the large domain, 132–163and 206–321 form the small domain and theremaining residues 164–205 form the clasp domainthrough the subunit interface between two anti-parallel  a -helices beneath a four-stranded inter-subunit anti-parallel  b -sheet. The large domain isconnected to the small domain by a flexible hingeregion.Figure1showsthefinalmodeloftheternarycomplex of   Ap IDH.An alignment based on secondary structuralassignment (Figure 2) revealed that most elementsof secondary structure were conserved withinsubfamily I. The superimposed C a traces showedthat the overall topology of   Ap IDH was almostidentical with that of   Ec IDH and  Bs IDH (Figure 3).The main differences are the extensions at bothtermini of   Ap IDH and the replacement of strand  K  and the preceding loop in the small domain of  Ec IDH by helix  g2  in  Ap IDH. In this region,  Bs IDHhas two helices,  g2  and  g3 , which are located in thevicinity of the active site of the neighbouringsubunit in the dimer, restricting the access to thephosphorylation site. 37 Strand  L  in this regionseems to be conserved in  Ap IDH,  Ec IDH and Bs IDH, but goes in the opposite direction in  Ap IDH and  Bs IDH compared to  Ec IDH.  Bs IDHhas an insertion of 13 residues in this region withrespect to  Ec IDH, whereas  Ap IDH has only onemoreresidue.TheRMSdifferencebetweenthelargedomain of native  Ap IDH  versus  that of   Ec IDH and Bs IDH was 1.49 A˚(using 211 C a atoms) and 1.74 A˚(211 C a atoms), respectively. For the small domaintogether with the clasp domain, the RMS difference between  Ap IDH  versus Ec IDH and  Bs IDH was3.28 A˚(188 C a atoms) and 1.43 A˚(192 C a atoms),respectively. The large difference between  Ap IDHand  Ec IDH is mainly due to the unique structuralfeature in the region of helix  g2  in  Ap IDH. Whenresidues 258–276 in this region were removed fromthe comparison, the RMS difference between thesmall and clasp domains of   Ap IDH  versus Ec IDHand  Bs IDH was only 0.82 A˚(171 C a atoms) and0.80 A˚(173 C a atoms), respectively.In total there are 15  a -helices in  Ap IDH (39.3%of the residues), of which 13 are conserved, and 15 b -strands (16.3%), of which all are conserved. Therearealsoseven3 10 helices(5.1%).Thetotalsecondarystructure content of 60.7% is only slightly higherthan that of   Ec IDH (57.2%) and reflects theformation of helices  g2 ,  n  and three additional 3 10 helices as well as a few loop deletions (described inthe thermostability section) in  Ap IDH. Structure and Thermal Stability of   A. pernix  IDH   561  Domain movements and differences in orientation between the subunits  Subunit B of the ternary complex had undergonea dramatic closure of the large domain, enclosingthe substrate and the cofactor in the active site,which is formed by the domain interface. Subunit Awas found still in an open conformation. Therelative difference in the orientation of the largedomain between subunit A and subunit B was 19 8 .The domain rotation axis in  Ap IDH was located inparallel with strand  E  and strand  F  (Figure 1a and b). It coincides with the domain definitions andapproximately with the pivot points of the domainmovements defined for  Ec IDH and  Bs IDH. 11 Thehinge loops in  Ec IDH defined by Doyle  et al . werenot adopted as hinge region. 38 The RMS difference between the small domainstogether with the clasp domains of the two subunitsof the ternary complex was 0.57 A˚for 190 C a atoms,whereas the RMS difference between the largedomains was 1.1 A˚using 215 C a atoms, indicatingdifferences other than just a rigid body movementof the large domain. Three major differences wereidentified when the domains of both subunits weresuperimposed; the C a traces of two NADP-bindingloops (residues 322–327 and 343–357) were trans-lated 1.6 A˚and 3.2 A˚, respectively, and helix  l (residues 396–400), which is also involved in NADP binding, was translated 2.3 A˚.Both subunits of native and pseudo-native  Ap IDH were found to be in an open conformation.The relative difference in the rotation of the largedomain between the subunits was 2–3 8  in eachdimer. In comparison, the difference between theclosed subunits of   Bs IDH is 2.9 8 . In  Ec IDH, thesubunits are related by a crystallographic 2-foldaxis, thus no differences between the subunits can be observed. However, two different crystal formsof   Ec IDH have revealed an open and a closedconformation with a relative diff erence in therotation of the large domain of 16 8 . 39 The confor-mation of native and pseudo-native  Ap IDH wassimilar to the open form of   Ec IDH (Figure 3).An investigation of the crystal contacts revealedwhy the conformations of the subunits were sodifferent in the ternary complex. Whereas the largedomain of subunit B almost had no crystal contacts,the large domain of subunit A was involved inseveral contacts that made the closure of the largedomain impossible upon substrate binding. It has Table 1.  Data collection and refinement statistics Native ApIDH Pseudo-native ApIDH Ternary ApIDHPDB code 1XGV 1TYO 1XKDWavelength (A˚) 0.996 0.9089 1.089Resolution limits (A˚) 2.2–39.8 (2.20–2.24) 2.1–39.9 (2.15–2.19) 2.3–39.7 (2.30–2.34)Mosaicity 0.5 0.5 0.6Unit cell parameters  a Z  b Z 107.57,  c Z 171.14 a Z b Z g Z 90 8 a Z  b Z 107.01,  c Z 179.82, a Z b Z g Z 90 8 a Z  b Z 107.57,  c Z 171.04, a Z b Z g Z 90 8 No. observed reflections 242,940 (10,181) 351,456 (13,556) 375,551 (16,016)No. unique reflections 51,805 (2522) 57,596 (2836) 45,289 (2259)Redundancy 4.7 (4.0) 6.1 (4.8) 8.3 (7.1)Completeness (%) 99.9 (100) 100 (100) 99.9 (100) I  / s  ( I  ) 17.97 (2.91) 17.62 (3.44) 25.42 (5.53) R merge  (%) 5.5 (49.7) 7.6 (52.6) 5.1 (36.7)Wilson  B -factor 40.1 30.3 45.7Space group  P 4 3 2 1 2  P 4 3 2 1 2  P 4 3 2 1 2 Refinement Non-hydrogen atoms 6689 6788 6789Non-hydrogen substrate atoms 0 32 (etheno-NADP) 122 (isocitrate, NADP)Non-hydrogen ion atoms 0 0 2 (Ca 2 C )Missing residues A1-5, B1-6, B114-9, B433-5 A1-5, A433-5, B1-6, B431-5 A1-4, A115-8, B1-5, B66-7B427-35Solvent molecules 161 200 195Resolution range (A˚) 2.20–39.8 (2.20–2.26) 2.15–39.9 (2.15–2.21) 2.30–35.0 (2.30–2.36) R cryst  overall (%) 22.5 (25.2) 22.3 (23.0) 22.6 (23.9) R free  (%) 25.1 (28.3) 24.9 (24.4) 24.8 (29.9)Ramachandran plot (excl Gly and Pro) in subunit A/BMost favourable region (%) 89.3/88.8 92.0 /91.1 90.0/86.7Allowed regions (%) 10.7/10.9 8.0 /8.9 9.7/13.3Disallowed regions (%) 0/0.3 0/0 0.3/0RMS deviation from ideal valuesBond lengths (A˚) 0.011 0.012 0.013Bond angles (deg.) 1.261 1.336 1.928Average  B -factors (A˚  2 )Total 19.5 20.8 27.3Isocitrate (A/B) 29.6/25.0NADP C (A/B) 31.7/31.0Etheno-NADP C 29.8Values in parentheses refer to data in the highest-resolution shell. 562  Structure and Thermal Stability of   A. pernix  IDH   Figure 1.  Ribbon representation of the ternary complex of   Ap IDH showing colour-coded domain definitions and theposition of the domain rotation axis (a and b) and a stereoview of an alignment of the open (green) and the closed (blue)subunits of the ternary complex of   Ap IDH (c). Subunit A was found in an open conformation, whereas subunit B wasclosed due to arotation ofthe large domain towards the small domain uponsubstrate binding. Thedifference in rotationof the large domain between the subunits was 19 8 . Structure and Thermal Stability of   A. pernix  IDH   563
Related Search
Similar documents
View more
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x